Scroll Top

Facebook Announces “Graph Search”

Facebook held a special event today, introducing a big new tool geared toward taming the massive amount of data it holds and providing a better way for customers to use that data. Before the event, titled ‘see what we’re building’, there was much speculation about what Facebook was going to be talking about — a new Facebook phone, a new mobile solution? When Mark Zuckerberg took the stage a few minutes after the 10am Pacific start time, we soon found that it was neither of these things, and possibly something larger.

“Today is about making new connections. Facebook is a community and a database of people’s lives.”

He started by reviewing the three pillars of the Facbook ecosystem, the first two of which you are already very familiar:

  1. Newsfeed – What’s going on with the people around me
  2. Timeline – Tell me something about this person
  3. Graph Search
The third pillar, which took center stage at today’s event, is The Social Graph – things your friends like, photos of friends doing things, places your friends have been. Facebook gathers an astounding amount of data and unique information about its users. Until now this data has been somewhat available to users, though not in any easily searchable or easily displayed way. So they’ve come up with Graph Search. Graph Search is essentially a way of taking all the social signals available in the Facebook system (e.g. what your friends like, where they’ve been, what others have done/liked), and presenting it to you as a simple search window, giving you deeper insight into the things that matter to those closest to you. Don’t think of this as a web search, rather a more personalized, natural, privacy aware, intuitive search.
“Friend connections, locations, likes, comments, tags, it all adds up to indexing all this content and making it so you can retrieve it instantaneously is a really hard technical problem.”
Taking a look at how this is all going to work, members of Mark’s team started by showing the search window with some basic queries: ‘music my friends like‘, ‘restaurants in chicago‘, ‘tv shows my friends like‘. Results you’ll see are based on the people you care most about. Beyond your friends’ results, data are then sorted by mutual friends and other Facebook signals.

Next they looked at how this search can be used for places. By using this tool you can search for places to go based on where your friends have traveled and what they liked. You can refine searches by type, liked by, location, and visited by my friends, similar to Foursquare.

Beyond looking at just those topics of interest to the people closest to you, this social search is also an amazing way to make new connections and tap into to the collective knowledge and experiences of friends-of-friends and beyond. For example, it is meant to help you find people you met through a mutual friend or at a party. This type of approach will also be useful for recruiting. For example, looking up ‘Friends of current employees‘ is a great place to start.

“Social graph provides answers, not links.”

Moving on next to pictures, they discussed the new ways you’ll be able to look up your friends photos. Find the best photos of your friends (based on likes, comments, etc), photos of friends taken at certain locations, photos of friends taken during a certain period of time, even simply ‘Photos I’ve Liked‘. This should reveal some interesting results. It’s like Google Image search, except it sources your friends’ public photos. An interesting spin in this category is rather than people, the option to look up places. For example, ‘Photos of Berlin.’ While this will only provide results for photos that are public, you are able to get insights about locations based on real-time user input.

The next area highlighted in the presentation was ‘Interests.’ Until now you had to go to someone’s timeline to find out your friends’ interests. With Graph Search, you’ll be able to query specific topics and find out more broadly, what your friends are interested in, what they like, and what they would recommend. For example, you can search ‘movies my friends like‘ or ‘TV shows my friends like.’ Taking this out to a higher level, you could look up ‘TV shows liked by doctors‘ or ‘music liked by people who like Barack Obama.’

Other ideas for searches that were highlighted on stage today:

  • ‘Dentists liked by my friends’ – list of dentists, maps, hours, addresses, and which friends like that person
  • ‘Restaurants liked by my Indian friends’ – list of top-rated Indian restaurants. Can also use tools on the right to narrow down places to one type of cuisine
  • ‘Bars in Dublin liked by people who live in Dublin’ – places the locals like
  • ‘Countries my friends have visited‘ – list countries visited, ranked by how many have been there. You can click a country to see which friends have been there.

With all of these new ways of finding information, they next talked about privacy. Looking at the new privacy shortcuts announced a few weeks ago they demonstrated how users can more easily see and adjust what can be seen by friends and what is public. It was also noted that users will have time to review their information prior to public launch.

“People are going to care about what shows up about them through search, so we have some great tools to answer that question.”

After the demonstrations, Mark Zuckerberg again took the stage to wrap things up. He emphasized again how early they are in the process of launching this and that it will be rolled out very slowly to allow users to prepare and review their information. He stated that this is not geared toward replacing web search and doesn’t think people will use it as such. “Graph Search is a really big project that will take years and years to index the whole map and the graph, but we’re really excited about people, photos, places and interests today.” The limited beta rolls out today and the public launch will take place over the next few weeks and months.


“We’re excited to get it into people’s hands. This is who we are. We love building things like this.”

Today’s event didn’t touch a couple of areas that people had thought it might, especially something mobile-related. Since mobile is such a huge concern and opportunity for Facebook, it will be interesting to see how Graph Search is developed for use on mobile devices. At first look it appears to be something users will have to use on their desktop, though as Mark suggested, they are very early in development. The potential that Graph Search has is quite daunting, especially if they can make the design more mobile-friendly and integrate some predictive behaviors that will help guide users toward things they might want to know based on things they’ve done, places they are going, etc. Graph Search takes Facebook a step closer to harnessing its massive database and making that data work for its users. As Mark stated, this is not meant to replace web search and won’t. After all, Google isn’t going anywhere. But this functionality certainly is very close to Google’s core business. Facebook has an immensely valuable database and each time they take a step toward better using that information, competitors like Google, LinkedIn, and Yelp need to pay very close attention. We’ll see what’s next!

Photos: The Verge

This article was written by Jeff Howland – Online Marketing Strategist at Dream Local Digital.

Privacy Preferences
When you visit our website, it may store information through your browser from specific services, usually in form of cookies. Here you can change your privacy preferences. Please note that blocking some types of cookies may impact your experience on our website and the services we offer.